Abstract
Dasheen mosaic virus (DsMV), Turnip mosaic virus (TuMV), Konjac mosaic virus
(KoMV) and Zantedeschia mild mosaic virus (ZaMMV) are important potyviruses previously identified in calla lily plants in Taiwan. In order to save time and cost of virus detection, a multiplex RT-PCR assay was developed for these calla potyviruses.
Specific primers for each virus were designed based on the sequences of 3’ terminal region of respective viruses. To prevent the false negative results, a primer pair specific to plant mitochondrial nad5 mRNA was used to produce a 185-bp fragment as an internal control of RT-PCR. The specificities of primers were confirmed by means of simplex and multiplex PCR assays. Optimal primer concentration ratio was identified by multiplex PCR assay. Total RNAs purified from virus-infected plants were used directly or mixed in different combinations, and then tested by multiplex RT-PCR. The result indicated that the expected RT-PCR products could be specifically amplified and identified on the basis of their molecular sizes. The detection sensitivity of multiplex RT-PCR was 25-625 times higher than that of indirect-ELISA (I-ELISA) depending on the virus. When applied to field surveys, multiplex RT-PCR could detect more single as well as mixed infection samples than I-ELISA. Accordingly, our multiplex RT-PCR assay provides a simple, rapid and reliable method for multiple potyvirus detection in calla lily.
Introduction
Calla lily is the common name of the members of Zantedeschia spp. which belong to the family Araceae and are classified into seven species and two subspecies (Kuehny 2000). As tropical plants native to Africa, calla lilies are now grown as outdoor garden plants, cut flowers and flowering potted plants. Many cultivars of calla lily have been introduced into Taiwan from New Zealand, the Netherlands and the United States for more than 10 years. However, virus diseases are one of the major factors limiting calla lily production (Chen et al. 2003; Huang et al. 2007). Calla lily has been reported as the natural host of various plant viruses, mainly potyviruses (Huang et al.
2007). There are six potyviruses identified in this plant and five of them have been found in Taiwan, including Calla lily latent virus (CLLV) (Chen et al. 2004), Dasheen
mosaic virus (DsMV) (Chen et al. 1998), Turnip mosaic virus (TuMV) (Chen et al.
2003), Zantedeschia mosaic virus (ZaMV) (Chang et al. 2001), and Zantedeschia mild
mosaic virus (ZaMMV) (Huang and Chang 2005). Among them, DsMV is a
well-known potyvirus and reported in many kinds of aroid plants (Rana et al. 1983;
Zettler and Hartman 1987). TuMV is an important cruciferous plant virus with a broad host range. ZaMV which was recently classified as the isolate of Konjak mosaic
virus (KoMV) is probably the most prevalent virus infecting calla lily (Chang et al.
2001; Kwon et al. 2002; Nishiguchi et al. 2006; Huang et al. 2007). A newly
identified calla virus, ZaMMV, was found widely spread in the fields in Taiwan and probably in other countries (Huang et al. 2007). The major symptoms of calla lily induced by these four viruses are mosaic, yellow stripe, green island and mild mosaic (Zettler and Hartman 1987, Chang et al. 2001; Pham et al. 2002; Chen et al. 2003;
Huang and Chang 2005). CLLV alone does not produce symptoms in calla lilies and may not have a direct impact on the crop (Chen et al. 2004). For that reason we selected DsMV, TuMV, KoMV (ZaMV) and ZaMMV as the detection targets in our study. According to the field surveys in Taiwan, mixed infections by potyviruses are very common in calla lilies (Huang et al. 2007). Severe or mixed infection could cause leaf and flower distortion, stunting, growth reduction and yield loss (Hu et al.
2007).
Tissue culture is an important propagation method for calla lily in addition to tuber production in Taiwan. Viruses can be maintained within the plants during the growing season and storage stage due to systemic viral infection. Therefore, tissue culture together with reliable virus detection methods is essential for production of virus-free plantlets and tubers. Several detection methods for individual calla potyviruses were recently developed such as reverse transcription-polymerase chain reaction (RT-PCR), immunocapture RT-PCR (IC-RT-PCR), dot-blot hybridization, and enzyme-linked immunosorbent assay (ELISA) (Huang et al. 2005; Hu et al. 2007;
Huang et al. 2007). These methods target only single viruses in one reaction. To save time and cost of virus detection, multiplex RT-PCR is chosen because it can rapidly detect multiple targets in one single assay with a small amount of sample. This technique has been successfully applied to many plants for virus detection, including apple (Menzel et al. 2002), citrus (Roy et al. 2005), olive (Bertolini et al. 2001), potato (Nie and Singh 2000; Du et al. 2006), orchids (Lee and Chang 2006), and other crops.
In this study, we established a multiplex RT-PCR system for simultaneous detection of four calla potyviruses in field samples. To rule out false negative results, one primer pair specific to plant mitochondrial NADH dehydrogenase (nad5) gene was incorporated to amplify the product of plant nad5 mRNA as the internal control.
This is the first multiplex RT-PCR developed for aroid plants.
Material and methods
Virus isolates and plant materials
Four potyviruses, DsMV, TuMV, ZaMMV and KoMV (ZaMV), were previously isolated from calla lilies. Isolates of DsMV-ZAN (Chen et al. 1998), KoMV (ZaMV-ZAN) (Chang et al. 2001), ZaMMV-ZUN (Huang and Chang 2005) and TuMV-ZAN were separately maintained on tissue culture plantlets of cultivar ‘Black Magic’ or Philodendron selloum by mechanical inoculation (Huang and Chang 2005).
TuMV was also maintained in Nicotiana benthamiana. These plants were kept at 25oC with 16 h photoperiod in a greenhouse. For the disease survey, field-grown calla lily plants were randomly collected from the Taiwan Seed Improvement and Propagation Station (TSIPS) in Taichung County.
Plant total RNA extraction
Plant tissues collected from healthy, virus-inoculated and field grown plants were used to extract total RNA following the protocol of Plant Total RNA Extraction Miniprep System (Viogene, Sunnyvale, CA, USA). In brief, 0.1 g leaf tissue was ground into fine powder in liquid nitrogen and then transferred to a microfuge tube.
After being mixed with 450 μl PRX extraction buffer, the lysate was filtered using a Shearing Tube. The filtrate was mixed with 230 μl absolute ethanol, transferred to a
new Plant Total RNA Mini Column, and filtered by centrifugation. This column was then washed once with WF Buffer and twice with WS Buffer. Finally, plant total RNA was eluted with 50 μl RNase-free ddH2O. The quality of total RNA was analyzed by 1% agarose gel electrophoresis. Plant total RNA was used directly for RT-PCR or stored at -20oC for further use.
Primer design
Specific primer pairs of four calla potyviruses were designed based on the sequences of 3’ terminal region of each individual virus, with the help of the Primer Premier 5 program (Premier Biosoft Int., Palo Alto, CA, USA) to avoid primer dimer formation.
The names, targets and sequences of primers and the expected product sizes are shown in Table 1.
Viral and plant nad5 cDNA clones construction
Plant mitochondrial nad5 gene (mt) and four viral cDNA fragments were prepared from virus-infected plant total RNA by RT-PCR amplification. The first-strand cDNA was synthesized using dT-Bam (5’-AGCTGGATCC(T)18-3’) or mtR1 primers. PCR reactions were carried out using dT-Bam and PNIbF1 (5’-GGBAAYAATAGTGGNCAACC-3’) primers for potyviruses (Hsu et al. 2005), or
mtR1/mtF2 primers for plant mitochondrial nad5 gene gene. RT-PCR products were analyzed in 1% agarose gels and the desired cDNA fragments were purified by GFXTM PCR DNA and Gel Band Purification Kit (Amersham Pharmacia Biotech, Piscataway, NJ, USA). The purified fragments were cloned into pGEM-T® Easy Vector (Promega, Madison, WI, USA). The correct cDNA clones were confirmed by sequencing and then used as templates for PCR experiments.
Multiplex PCR
Five pairs of reverse and forward primers (Table 1) were used in the multiplex PCR reaction. Different primer mixtures containing each primer 10x concentrated were prepared according to the experimental design. The final concentration of every primer was 0.125 μM in original multiplex PCR. To adjust the primer ratio, TuMV primer concentration was decreased to 0.5 and 0.25x and DsMV primer was increased to 1.5x of original concentration (0.125 μM). For multiplex PCR, the 20 μl reaction mix contained 2 μl template mixture (2 ng per cDNA clone), 2 μl 10x primer mixture, 2 μl 10x DyNAzymeTM II DNA polymerase buffer (Finnzymes, Inc., Espoo, Finland), 0.5 μl dNTPs (10 mM), 0.5 μl DyNAzymeTM II DNA polymerase (2 U μl-1, Finnzymes, Inc.) and 13 μl ddH2O. The amplification was carried out in GeneAmp® PCR system 2400 or 2700 (Perkin-Elmer Applied Biosystems, Foster City, CA, USA).
PCR program for DNA synthesis was 95oC for 5 min, followed by 30 cycles of 95oC for 35 s, 56oC for 35 s, 72oC for 1 min 30 s, and a final elongation step at 72oC for 7 min. Eight μl of PCR products were analyzed by 1.5% agarose gel electrophoresis in 1X TAE buffer (40 mM Tris-acetate, 1 mM EDTA). The DNA fragments were stained with ethidium bromide for 10 min and examined under UV illumination.
Multiplex RT-PCR
The multiplex RT-PCR protocol has separate RT and PCR steps. Total RNA was extracted as described above. For 25 μl RT reaction, 7 μl plant total RNA together with 2 μl dT-Bam primer (10 μM) and 1 μl mtR1 primer (5 μM) were heated at 65oC for 10 min, cooled at 4oC for 5 min and then 15 μl RT mixture [5 μl 5x first strand buffer (Promega, Madison, WI, USA), 1 μl dNTPs (10 mM), 0.5 μl rRNasin (40 U μl-1, Promega), 1 μl AMV reverse transcriptase (10 U μl-1, Promega) and 7.5 μl ddH2O] was added. After incubating RT reaction solution at 42oC for 60 min, the multiplex PCR reaction was performed as previously described except that 2 μl RT product was used as template.
Indirect enzyme-linked immunosorbent assay (I-ELISA)
The antisera to DsMV, TuMV, ZaMMV and KoMV were previously prepared in our
laboratory using recombinant capsid proteins as the antigens (Huang et al. 2005; Hu et al. 2007; Huang et al. 2007). I-ELISA was performed according to the protocol of Agdia Inc. (Elkhart, IN, USA) with some modification. One hundred mg of plant tissue was ground in 1 ml of indirect sample extraction buffer [ISE buffer, 15 mM Na2CO3, 35 mM NaHCO3, 2% polyvinylpyrrolidone (MW 40,000), pH 9.6]. One hundred μl of the extracts were coated to the 96-well ELISA plate and then assayed as previously described (Lee and Chang 2008). Each sample assayed in triplicate. A sample was regarded as positive if the A405 value exceeded twice the mean value of healthy controls.
Results
Specificity of detection primers tested by simplex and multiplex PCR
The performances of the designed primers were first tested using viral and plant nad5 cDNA clones as templates in simplex and multiplex PCR assays. The final concentration of each primer was 0.125 μM in these PCR reactions. Initially, the specificity of individual primer pairs was analyzed by PCR reaction with single cDNA template. The results of simplex PCR demonstrated that the expected PCR products were successfully generated by each single specific primer pair (Fig. 1a). When individual cDNA clones were tested by the multiplex PCR reaction, the expected fragments were specifically amplified by the primer mixture of TuMV, ZaMMV, DsMV, KoMV and mt control (Fig. 1b). Although a lower yield of the products of DsMV, KoMV and mt were obtained in multiplex PCR compared with simplex PCR, all tested primers were specific to their targets. To further test the specificity of these primers with five templates simultaneously, the same amount of each individual cDNA clone was combined and then used for the multiplex PCR assay. According to the gel analysis, the PCR products of TuMV, ZaMMV, KoMV and mt control were successfully amplified but the DsMV fragment was hardly visible (Fig. 1c) indicating that the multiplex PCR needed modification in order to detect multiple targets.
Optimization of multiplex PCR reaction
To optimize the multiplex PCR for detection of multiple targets, different concentration ratios of primer pairs were tested for amplification efficiency. In previous PCR assays, the same final concentration (0.125 μM) of each primer was used and thus 0.125 μM was assigned as ratio 1x in the primer ratio test. From our preliminary tests, DsMV amplification was hampered when using the same amount of TuMV and DsMV primers (data not shown). Consequently, five primer sets (I-V) with different primer ratios were prepared by decreasing TuMV primer ratio from 1 to 0.5 and 0.25x, and at the same time increasing DsMV primer ratio to 1.5x (Fig. 2). At first individual viral and mt control templates were each tested by five sets of primer mixture, and all expected fragments were produced in multiplex PCR (Fig. 2a-d).
Although the size of TuMV amplicon was the largest among five amplified targets, TuMV PCR product yield was reduced only slightly when its primer ratio was lowered to 0.25x (Fig. 2a, lanes 3 and 5). On the other hand, DsMV products increased when the primer ratio was raised to 1.5x (Fig. 2c, lanes 4 and 5).
Furthermore, we tested the amplification efficiency of these primer mixtures with five cDNA templates simultaneously. The result clearly indicated that reducing the concentration of TuMV primer to ratio 0.25x allowed a significant increase in DsMV amplification rate and raising the DsMV primer ratio had a similar effect (Fig. 2e,
lanes 3-5). Since primer sets III to V were able to amplify all five targets in multiplex PCR reaction, we chose the primer set V for subsequent RT-PCR experiments.
Specificity of multiplex RT-PCR against total RNAs from virus-infected plants
Purified total RNAs from calla lilies individually infected with TuMV, ZaMMV, DsMV and KoMV were used for cDNA synthesis using dT-Bam and mtR1 primers and multiplex PCR was performed with the primer set V. The correct virus targets were specifically amplified and the mt control fragment was consistently generated in all virus-infected samples (Fig. 3a). However, field-grown calla lilies are frequently infected by mixtures of two or more potyviruses. To test the ability to detect multiple viruses, artificial mixed-infection samples prepared by mixing two, three, or four different virus-infected plant total RNAs were assayed by multiplex RT-PCR. The expected RT-PCR products including virus and mt targets were correctly amplified in all tested virus combinations (Fig. 3b).
Comparison of detection sensitivity between multiplex RT-PCR and I-ELISA
To compare the detection sensitivity, the same amount of virus-infected calla lilies (0.1 g) were used to purify plant total RNA and to prepare original plant extract with 10 volume of ISE buffer. Afterward fivefold serially diluted RNA and sap samples
were prepared with healthy plant total RNA and extract as diluents, and then tested by multiplex RT-PCR and I-ELISA, respectively. The highest dilution at which multiplex RT-PCR showed positive result was 5-3 for TuMV, 5-4 for DsMV and KoMV, and 5-5 for ZaMMV. However, the detection limit of I-ELISA assay was determined as 50 for KoMV, and 5-1 for TuMV, ZaMMV and DsMV (Fig. 4). Although KoMV had the lowest ELISA value in this test, the detecting limitation of multiplex RT-PCR still reached to 5-4 dilution (Fig. 4d). In summary, the detection sensitivity of multiplex RT-PCR was 25-625 times higher than that of I-ELISA depending on the virus.
Although a prozone effect in I-ELISA was noted for TuMV and ZaMMV when 1:10 (w/v) extracts were used compared with 1:100 (w/v) as suggested by AGDIA, use of the former had minimal effect on the results of final assay comparisons (data not shown).
Detection of calla potyviruses in field samples by multiplex RT-PCR and
I-ELISA
Fifty full-expanded leaves of calla lily were randomly collected from the field of TSIPS in Taichung County. These samples were used for the detection of calla potyviruses detection by multiplex RT-PCR and I-ELISA assays at the same time.
According to the data, 72% (36/50) of calla lily samples tested positive for virus
infection by multiplex RT-PCR, and about two thirds (23/36) of infected samples were mixed infections. By contrast, only 50% of the samples were positive in potyvirus-specific I-ELISA (Table 2). KoMV, found in 68% (34/50) and 44% (22/50) of the samples tested by multiplex RT-PCR and I-ELISA, was the dominant virus in the field. ZaMMV was identified in 44% (22/50) and TuMV was detected in 22%
(11/50) of calla lily plants by multiplex RT-PCR. The previously important aroid plant virus, DsMV, was only detected in one sample using multiplex RT-PCR (Table 2). It is apparent that the multiplex RT-PCR method had better sensitivity than I-ELISA to detect calla potyviruses in field samples.
Discussion
This paper describes a multiplex RT-PCR method for simultaneous detection of four potyviruses in calla lily with coamplification of a plant internal control. We selected DsMV, TuMV, ZaMMV and KoMV as the viral targets because these four viruses caused obvious symptoms and yield loss in calla lilies. In multiplex PCR, a DsMV fragment was barely amplified from a multiple-template mix when the final concentration of individual primer was equivalent (Fig. 1c). For unknown reasons, the TuMV primers seemed to interfere with the amplification of DsMV. We solved the problem by increasing the DsMV primer concentrations and decreasing the TuMV primer concentrations as well (Fig. 2). Although several studies indicated that short fragments are amplified more efficiently than longer fragments (Du et al. 2006; Hu et al. 2007), the amplification efficiency of the TuMV fragment was as good as other viral targets in all mixed samples (Figs. 2e and 3b). It might be due to the Tm value of TuMV primers (64 and 68oC) being higher than those of other viral primers, and hence more effectively amplifying the viral target during the PCR process. Even though short fragments were usually selected for multiple target detection, fragment size up to 814 and 942 bp were successfully used for virus detection in citrus trees (Roy et al. 2005). Consequently, large target fragments can still be utilized in multiplex system if the primer design and the PCR condition are optimized.
To avoid the false negative results, different plant internal controls have been introduced into virus detection systems, such as 18S rRNA (Du et al. 2006), chloroplast NADH dehydrogenase ND2 subunit (ndhB) mRNA (Thompson et al.
2003) and mitochondrial NADH dehydrogenase (nad5 and nad2) mRNAs (Menzel et al. 2002; Du et al. 2006). Our previously designed mt primers, mtF2 and mtR1, consistently amplified a 185-bp cDNA fragment of nad5 mRNA from orchid total RNA (Lee and Chang 2006). When the same primers were applied to calla lily, the amplified internal control fragments indicated the success of total RNA extraction and multiplex RT-PCR process (Fig. 3). Nevertheless, the consequences of our preliminary tests showed that the amplification efficiency of the nad5 fragment was greater than that of the potyvirus targets and sometimes interfered with virus amplification. Therefore, the mt primer was reduced to 1/4 amount of oligo(dT) primer in RT step to lower first-strand cDNA synthesis of nad5 mRNA; and then the mt primer concentration was also adjusted from 0.25 μM to 0.125 μM in PCR step to improve the results (data not shown). In our opinion, the optimal primer concentration for a plant internal control is influenced by the features of plant species and viral targets, and needs adjustment in every virus detection system.
Due to the characteristic poly(A) tail at the 3’ end of potyvirus genome, we chose an oligo(dT) primer to synthesize first-strand cDNA of potyviruses.
Nevertheless, the mtR1 primer was used in addition to the oligo(dT) primer in the RT step to ensure sufficient cDNA synthesis of the plant internal control. In our system, utilization of a universal oligo(dT) primer in the RT step lowered the primer complexity and the PCR conditions could be easily adjusted to facilitate the detection of virus containing a 3’-poly(A) tail. Addition of specific primer pairs could expand the assay to detect calla viruses with and without 3'-poly(A) tail as previously reported in potato virus detection (Nie and Singh 2000).
The detection limit of our multiplex RT-PCR was higher than I-ELISA according to the sensitivity comparison experiment. This was again confirmed by field disease survey, since multiplex RT-PCR identified more single as well as mixed infection samples than I-ELISA. From 50 calla lily plants randomly collected from the field, 36 samples tested positive for potyviruses by multiplex RT-PCR. KoMV and ZaMMV were the two dominant viruses found, in 68% and 44% of the tested samples respectively, and most of them from mixed infections. These data agreed with our investigation in the field survey during 2003-2004 (Huang et al. 2007). According to the prior report, field grown calla lily might be infected by TuMV through random
The detection limit of our multiplex RT-PCR was higher than I-ELISA according to the sensitivity comparison experiment. This was again confirmed by field disease survey, since multiplex RT-PCR identified more single as well as mixed infection samples than I-ELISA. From 50 calla lily plants randomly collected from the field, 36 samples tested positive for potyviruses by multiplex RT-PCR. KoMV and ZaMMV were the two dominant viruses found, in 68% and 44% of the tested samples respectively, and most of them from mixed infections. These data agreed with our investigation in the field survey during 2003-2004 (Huang et al. 2007). According to the prior report, field grown calla lily might be infected by TuMV through random